Salesforce Survives AI, but Its Clients No Longer Want the Same
Some companies don’t die from a single blow. They fade into irrelevance while their financial statements still look respectable. Salesforce, having dominated corporate CRM for 27 years, now finds itself at a tipping point where analysts’ headlines and organic data tell different stories.
The stock has plummeted 35% in the past year, trading around $170. Simultaneously, Agentforce, its AI investment, generated $800 million in its most recent quarter. Citizens maintains a target price of $315 with a buy recommendation, while Northland dropped it to $229. This type of divergence among analysts isn’t just noise; it signals that no one knows exactly what the market will ask from Salesforce in 18 months.
What we do know is that the organic growth of current remaining performance obligation (cRPO) hovers around 9% when adjusted for the acquisition of Informatica and exchange rates. This figure falls below the 10-12% threshold that analysts consider necessary to justify a high-growth platform narrative. The total cRPO of $35.1 billion looks better on paper, but paper is precisely where that figure lives.
The Issue Isn’t AI, It’s Who’s Charging for It
Morgan DeBaun, CEO of Blavity, claimed that switching from Salesforce CRM to AI alternatives before early 2027 could represent cost savings of 50% to 60%. That statement carries more weight than any analyst projection because it comes from someone who pays the monthly bill.
What DeBaun is describing isn’t dissatisfaction with the technology itself—it's dissatisfaction with the technology’s pricing model. For years, Salesforce built a scheme where additional features were monetized as add-ons. This works perfectly when a customer has no alternative. It falters when competitors like Anthropic with Cowork and various AI-native platforms provide intelligence in the base rate.
Here’s the invisible mechanics that matter: Salesforce overserved the enterprise segment with layers of customization, integrations, and functionalities that most of its medium-sized customers never utilized to the fullest. This complexity justified premium pricing. Now, that same complexity creates friction when the market demands agility to deploy autonomous agents without relying on a team of certified consultants.
Medium and large companies are hiring for different needs than what Salesforce has been selling for two decades. Previously, they sought relationship management: customer databases, sales pipelines, reporting. Now, they seek operational autonomy: software that acts, not just records. This is where Salesforce's historical value proposition collides with its own legacy architecture.
Agentforce and the Risk of Building on Costly Foundations
The $800 million quarterly revenue from Agentforce is real and relevant. The 5% stock price jump following that results announcement reflects market celebration of the signal. However, there’s a structural question that this revenue still doesn’t address: Agentforce coexists with the existing Salesforce platform; it does not replace it.
This matters because the proposition from native AI competitors is exactly the opposite: start fresh, without the technical debt of decades of traditional CRM, and build workflows around the agent, not around the record.
The accelerated buyback of $25 billion in shares and personal stake acquisitions by board members in $500,000 blocks signal internal confidence. The $15 billion investment in San Francisco, which includes an AI incubation center, indicates the company is betting on building its own talent hub. These are both defensive and offensive moves, consistent with a company that knows it has between 12 to 24 months to reallocate its narrative.
Salesforce’s internal study among Chief Information Officers shows a 282% adoption rate for AI and that 77% of workers expect to eventually rely on AI for autonomous operations. This data is useful for sales decks, but it also reveals something more uncomfortable: if Salesforce’s own clients are adopting AI at that pace, the pressure for providers to deliver autonomous results is already upon them.
What This Case Informs Any Software Selling Company
The pattern I observe in Salesforce isn’t that of a declining company. It's the behavior of a company that built its competitive advantage on accumulating customer data and workflows—a type of lock-in that was nearly impenetrable for two decades—but now finds that AI has lowered the cost of migration.
When switching CRMs cost months of consulting, retraining, and risks of losing historical data, Salesforce’s price was one of certainty. When an AI-native platform promises to import history, map workflows in weeks, and charge less from day one, that certainty is no longer exclusive.
For any company operating software as a service with medium customers—this directly applies to smaller companies competing in vertical markets—this case serves as a concrete operational warning: the moment your client starts calculating the cost of leaving, you've already lost part of the renewal conversation. DeBaun’s suggestion to switch providers wasn’t an emotional reaction; it stemmed from a financial projection of 50-60% savings. That is measured in a spreadsheet, not in a satisfaction survey.
The lesson for a medium-sized company selling technology is that keeping the client within the platform due to operational friction is a strategy that works until someone reduces that friction to zero. At that point, the only argument left is the results the product delivers. Not the integrations. Not the years of history. The outcome.
The work companies were hiring Salesforce for was never just to manage contacts or automate emails: it was to reduce business uncertainty by having total visibility over the pipeline. Agentforce may deliver that autonomously, but if it does so at a price that doesn’t compete with simpler platforms, Salesforce will have solved the technological problem and still lost the customer.
